C-Urea Breath Test in Helicobacter pylori Diagnosis ...

3 downloads 0 Views 806KB Size Report
Jul 8, 2009 - patients was initially corroborated by histology (Warthin-. Starry staining). Histology. For histologic examination, formalin-fixed biopsy samples.
Scandinavian Journal of Gastroenterology

ISSN: 0036-5521 (Print) 1502-7708 (Online) Journal homepage: http://www.tandfonline.com/loi/igas20

13

C-Urea Breath Test in Helicobacter pylori Diagnosis and Eradication Correlation to Histology, Origin of ‘False’ Results, and Influence of Food Intake H. J. Epple, F. W. Kirstein, C. Bojarski, J. Frege, M. Fromm, E. O. Riecken & J. D. Schulzke To cite this article: H. J. Epple, F. W. Kirstein, C. Bojarski, J. Frege, M. Fromm, E. O. Riecken 13

& J. D. Schulzke (1997) C-Urea Breath Test in Helicobacter pylori Diagnosis and Eradication Correlation to Histology, Origin of ‘False’ Results, and Influence of Food Intake, Scandinavian Journal of Gastroenterology, 32:4, 308-314, DOI: 10.3109/00365529709007677 To link to this article: http://dx.doi.org/10.3109/00365529709007677

Published online: 08 Jul 2009.

Submit your article to this journal

Article views: 51

View related articles

Citing articles: 7 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=igas20 Download by: [FU Berlin]

Date: 09 April 2017, At: 22:24

13C-Urea Breath Test in Helicobacter pylori Diagnosis and Eradication Correlation to Histology, Origin of ‘False’ Results, and Influence of Food Intake H. J. EPPLE, F. W. KIRSTEIN, C. BOJARSKI, J. FREGE, M. FROMM, E. 0. RIECKEN & J. D. SCHULZKE

Depts. of Gastroenterology, Clinical Physiology, and Pathology, Universitatsklinikum Benjamin Franklin, Freie Universitat, Berlin, Germany Epple HJ, Kirstein FW, Bojarski C, Frege J, Fromm M, Riecken EO, Schulzke JD. ”C-Urea breath test in Helicobacter pylori diagnosis and eradication. Correlation to histology, origin of ‘false’ results, and influence of food intake. Scand J Gastroenterol 1997;32:308-314. Background: Which protocol is optimal for the 13C-urea breath test (UBT) for Helicobacter pylori detection is controversial. This study aimed to characterize a very simple UBT protocol for the clinical routine (two-point-analysis performed with 75 mg “C-urea and citric acid) with special consideration of ‘false’ UBT results. Results: UBT was evaluated in reference to histology (Warthin-Starry). In mismatching results re-gastroscopy was performed. By UBT, 74 of 77 patients with H. pylori-positive histology were detected (sensitivity, 96%). The false-negative UBTs were due to low colonization densities during spontaneous H. pylori elimination or pyloric obstruction. Seven of 49 patients with negative histology had a positive UBT, but re-gastroscopy showed that all of them had a positive histology when multiple antral biopsy specimens were taken (UBT specificity, 100%). UBT correlated only weakly with H. pylori colonization density. No correlation was found between UBT and gastric neutrophil and lymphocyte infiltration. UBT reproducibility was excellent (93 of 94 in a 6-month period). Non-fasting conditions induced a shift to lower UBT results in H. pylori-positive and to higher UBT results in negative patients, resulting in 2 of 10 false-positive and 1 of 10 false-negative UBTs. Conclusion: This simple version of the urea breath test combines the highest sensitivity and specificity with excellent reproducibility. It is superior to histologic detection of H. pylori in the clinical routine and an optimal tool for monitoring H. pylori eradication. Fasting conditions are required for the test. Key words: Fasting condition; gastritis; Helicobacter pylori; stomach; urea breath test Jorg-Dieter Schulzke, M.D., Medizinische Klinik und Poliklinik, Abteilung fur Gastroenterologie, UniversitatsklinikumBenjamin Franklin, Freie Universitat Berlin, D-12200 Berlin, Germany (fax: + 49 30 8445-4239)

Helicobacterpylori has been established as an etiologic factor in type-B chronic gastritis, peptic ulcer disease, and gastric carcinoma and lymphoma. H. pylori eradication has therefore become the main principle in the therapy of ulcer disease, and the role of this eradication in the treatment of gastric lymphoma (1) and H. pyfori-positive non-ulcer dyspepsia is still extensively studied (2). Several invasive and non-invasive diagnostic methods have been developed to detect H. pylori and to monitor eradication therapy. Most authors propose invasive-that is, endoscopydependent-methods for the initial diagnosis of H. pylori infection, because endoscopic examination of the upper gastrointestinal tract with histologic investigation of mucosal biopsy specimens yields additional, often essential, information. Indications for non-invasive detection methods beyond epidemiologic studies are i) control of eradication therapy, and ii) when the invasive H. pylori test result is negative but one still suspects H. pylori infection (‘sampling error’). The 13C-urea breath test (UBT) is the most important non-invasive diagnostic tool for H. pylori infection and has been successfully applied for all of these indications. The

UBT was originally restricted to clinical trials, but simplifications and lower costs due to smaller tracer amounts and reduced prices for mass spectrometers will make it available also for clinical routine examinations. The UBT for H. pylori detection was introduced in 1987 by Graham et ai. (3). In the original description -350 mg of 13Clabeled urea (5mg/kg) were used, and five breath samples obtained at 10-min intervals during 60 min after ingestion of the tracer were analyzed. Since then, many modifications of the UBT have been reported in the search for the least expensive and most convenient test protocol without impairment of the UBT’s high accuracy and reliability. In a recent study the minimum requirements for accurate H. pyZori detection by the UBT were defined (4). Analysis of only two breath samples, one taken before and the other 30 min after ingestion of the labeled urea, gave optimal results in terms of sensitivity, specificity, and positive predictive value. Citric acid instead of a fatty meal was used to inhibit gastric emptying in the UBT protocol of Eggers et al. (5). In their study a very small amount of ‘3C-labeled urea (75 mg) and a simple breath sampling procedure with four-point-analysis of

Urea Breath Test in Helicobacter Diagnosis

exhaled breath were used. The results of the UBT were normalized to endogenous COz production and showed a clear bimodal distribution for H. pylori-positive and -negative patients. The independence of UBT results from factors influencing endogenous COz production (for example, height and weight) was corroborated by another study (6). In the present paper a very simple UBT protocol combining a reduced tracer amount (75mg I3C-urea), a two-pointanalysis, and the use of citric acid was systematically investigated in the clinical routine. Special attention was focused on breath test results that differed from the respective histology results. The origin of these ‘false’ UBT results were evaluated in detail by re-gastroscopy. Furthermore, UBT results were compared with H. pylori colonization density and with the degree of mucosal leukocyte infiltration, and UBT reproducibility was examined by intraindividual re-testing. Finally, the influence of food intake was investigated, to further define technical details with this UBT protocol. SUBJECTS AND METHODS

309

of gastritis were graded in accordance with a modification of the Sydney classification at 400-fold magnification on two antrum and two corpus mucosa specimens (7). Colonization density was classified on Warthin-Starry-stained sections as absent (grade 0), focally few H. pylori (grade l), moderately dense (grade 2), and dense (grade 3). The activity of gastritis (density of neutrophil infiltration) and degree of gastritis (density of lymphocyte and plasma cell infiltration) were classified on HE-stained sections. The degree of neutrophil infiltration was classified as absent (grade 0), isolated neutrophils (grade l), and moderate (grade 2) and dense (grade 3) neutrophil infiltration of the lamina propria. Lymphocyte and plasma cell infiltration density was classified as absent (grade 0), uniform slight infiltration (grade l), moderate (grade 2), and dense infiltration (grade 3).

13C-Urea breath test The standard 13C-ureabreath test for detection of H. pylori was performed after an overnight fast. Patients drank 150 ml 0.1 N citric acid to delay gastric emptying. One minute later 75 mg 13C-urea (13C-urea, 99% purity, Cambridge Isotopes, Andover, N.H., USA) dissolved in 20 mlO.1 N citric acid was swallowed. Finally, 30 mlO.1 N citric acid was drunk to rinse the tracer from the oral cavity, pharynx, and esophagus. Exhaled breath samples were taken in duplicate before and 30 min after ingestion of the tracer in 10-ml glass vacutainers (Becton Dickinson, N.J., USA) by means of a straw. The I3CO2 to ‘2COz isotope ratio in the breath samples was measured with isotope ratio mass spectrometry (Optima, Fisons Instruments, Middlewich, England), and the relative abundance of ‘-?COZwas determined as 6 value (%o) versus PDB (Pee Dee Belemnite, international limestone standard (8)). The increase in the 6 value after ingestion of the tracer (66 value) was determined by subtraction: 66 (7%) = 6 value of base-line breath sample minus 6 value of breath sample 30 min after ingestion of the tracer. The 13C-urea breath test was considered positive when a 66 value was more than mean + 2 standard deviations (s) of 66 values of 42 H. pylori-negative patients (see Results). To investigate the influence of food intake on UBT results, 10 H. pylori-negative and 10 positive patients were tested twice. The tests were performed after a 12-h (overnight) fasting period and 30 min after lunch.

Patient recruitment H. pylori status was evaluated in 126 consecutive patients (56 male, 70 female; median age, 48 years (range, 17-84 years)) who attended our endoscopy department for routine endoscopy of the upper gastrointestinal tract. Two antral and two corpus forceps biopsy specimens were taken for histologic examination. The 13C-urea breath test was performed within 3 days of the endoscopy. Patients with gastric carcinoma or previous gastric surgery were excluded. Patients gave their informed consent, and the study was approved by the local ethics committee. To test the reproducibility of the UBT, 94 patients were retested 6 months after the first UBT. This included investigation of 10 H. pylori-negative volunteers and 14 H. pyloripositive and 70 -negative patients who had taken part in a randomized prospective eradication study and a study to evaluate the role of H. pylori in non-ulcer dyspepsia. Eradication regimens included bismuth subsalicylate, 1.8 giday ; tetracycline-HC1, 2 giday; and metronidazole, 1.2 giday; for 10 days (Australian triple therapy); amoxicillin, 3 gfday, plus omeprazole, 40 mgiday, for 14 days (German dual therapy); clarithromycin, 1&day, plus omeprazole, 40 mdday, for 14 days (English dual therapy) or tetracycline-HCI, 2 @day, plus omeprazole, 40mglday. The H . pylori status in all of these patients was initially corroborated by histology (WarthinStarry staining). Statistics The Kruskal-Wallis test was used to test for significant Histology differences in 66values of UBTs in the histologically defined For histologic examination, formalin-fixed biopsy samples subgroups. If a significant difference was found, Spearman were embedded in paraffin, and 5-pm-thin sections were rank correlation was used to determine the correlation stained with hematoxylin-eosin (HE) and Warthin-Starry. coefficient. This was performed in the Dept. of Pathology without Paired Student’s t test was used for comparison of 13C-urea knowledge of the UBT result. breath test results during fasting and non-fasting test conH. pylori colonization density and the degree and activity ditions; P < 0.05 was considered significant.

H. J. Epple et al.

01 I100 2O4

40

-

30

-

0

20

-

10-

0-

absent

0

07

3 4j

0

0

0

24

0

histology negative

(n=49)

-

0

histology positive (n=77)

Fig. 1. Results of the 13C-urea breath test (UBT) in a consecutive group of patients undergoing gastroscopy in our department. Patients were subdivided into 2 groups, with negative (n = 49) and with positive (n = 77) histology (WarthinStarry staining). The results of the UBT in these patients are given as 66 values (%o) as described in Methods. The normal range is indicated by the scattered line at 1.3%0.UBT results that did not match with the respective histology result are indicated (nos. 1-10) (see Results).

RESULTS Normal range The H . pylori-negative group ( n = 42) showed a mean 66 value of 0.40%0with a standard deviation of 0.44. Thus, the normal range (mean 2 s) amounted to